Dec 312004

While they were mentioned on Slashdot over two years ago, these “virtual keyboards” seemed to always be just a little bit further on the horizon, and to be incorporated into cell phones and PDAs, not as a stand-alone gadget amenable to hacking.

But it looks like one company, VKB has in production a stand-alone version with a serial port output. And at least one company in the US, “PC-notetaker” is selling them. For $199.

That’s just a bit expensive for casually buying one to hack apart and maybe turn into an interesting music performance gizmo. Maybe I’ll try one out anyway. I wonder how many simultaneous keypresses it can handle?

(oh wait, it looks like expansys has some for $160.95)

 Posted by at 12:31 am
Jun 152004

Andrew Choi has an OS X Programming blog which contains a lot of useful and interesting observations and test code as he builds a auto-accompaniment application.

Potentially useful to my KCS remake idea (assuming I can use MusicPlayer and not have to write my own scheduler) are his experiments with the MusicSequence and MusicPlayer APIs that are part of Core Audio (and not, confusingly, really considered part of Core MIDI). Particularly useful is his MusicSequenceTest code that shows how to use it in a lot clearer way than the official example given in /Developer/Examples/CoreAudio/Services/PlaySequence.

Aside: just found this page on with data flow diagrams on MIDI Operations in Core Audio that seems kinda useful.

Question: since there is a one-to-one relationship between MusicSequence and MusicPlayer, and the MusicPlayer is what has the transport controls (stop,start,etc), can I instantiate many MusicPlayers (say 128 of them) and start/stop them all at various with impunity, with multiple MusicPlayers running concurrently? If so, the hard part of KCS is done.

 Posted by at 3:36 pm